DQ-1220 Leak Questions

A candidate key is a minimal super key. A super key is any set of attributes sufficient to uniquely identify an entity instance, but it may contain attributes that are unnecessary for uniqueness. A candidate key removes that redundancy: if any attribute is removed from the candidate key, the remaining attributes no longer uniquely identify the entity.

DAMA-DMBOK2 makes this distinction explicitly: a candidate key is a minimal set of one or more attributes identifying an entity instance, and “minimal” means that no subset of the candidate key can perform the same unique-identification function.

Option B is incomplete because a set can uniquely identify an entity while still containing unnecessary attributes; that would qualify as a super key but not necessarily a minimal super key. A surrogate key is different: it is an artificial identifier introduced primarily for technical identification. Foreign-key composition and physical index structures likewise do not define candidate-key minimality.

This concept has direct Data Quality implications. Correctly defined candidate and primary keys support uniqueness and integrity, prevent duplicate entity instances, enable reliable referential relationships, and improve entity matching within Master Data Management. Key definitions should also be captured as structural metadata so profiling and quality rules can consistently test duplicate and orphan conditions.

Reference Topics: DAMA-DMBOK2 Chapter 5 — Data Modeling and Design; Keys; Candidate Keys; Chapter 13 — Uniqueness and Integrity; Metadata Management; Master Data Management.

===============

DAMA-DMBOK2 identifies four principal types of data that may be represented in data models: Category Information, Resource Information, Business Event Information, and Detail Transaction Information. Because options A, B, C and E are all legitimate categories, All of these is correct.

Category information classifies things—for example, classifying customers by market segment or products by size or colour. Resource information represents business resources required for operations, including entities such as Customer, Product, Supplier, Facility and Account. These resource entities frequently overlap strongly with Master and Reference Data Management.

Business event information records events created during operational processes, such as orders, invoices or withdrawals. Detail transaction information represents highly granular observations, including point-of-sale records, clickstream data, social-media interactions, sensor outputs and other high-volume events.

This classification matters to Data Quality because quality controls must reflect the semantics and lifecycle of the data being evaluated. Reference and master data require strong uniqueness, definition and consistency controls; event data require integrity and timeliness; high-volume detailed data often require automated profiling and statistical monitoring.

Data models also generate important metadata—definitions, relationships, domains and constraints—which becomes the basis for measurable Data Quality rules.

Reference Topics: DAMA-DMBOK2 Chapter 5 — Types of Data that are Modeled; Chapter 10 — Reference and Master Data; Chapter 13 — Data Quality Rules and Dimensions; Metadata Management.

===============

The definition describes Data Policies. DAMA-DMBOK2 states that data policies are directives that translate organizational principles and management intent into fundamental rules controlling how data is created, acquired, protected, maintained, assured for quality, and used.

The distinction between governance and policy is important. Data Governance is the broader authority and decision-making framework. Policies are one of the principal instruments produced or sponsored through that framework. They define what must or must not occur. Standards and procedures subsequently translate those policies into specific implementation requirements and operational methods.

For Data Quality, a policy might require business-critical data to have designated ownership, documented definitions, measurable quality rules, monitored thresholds, and formal remediation processes. The detailed thresholds themselves may reside in standards or rule repositories rather than in the high-level policy.

Metadata Management makes policy operational by documenting definitions, classifications, ownership, lineage and applicable quality rules. Master Data Management applies those policies to shared business entities and controlled reference values. Data Stewards then monitor compliance and escalate violations through governance mechanisms.

Data asset valuation is concerned with determining economic value, while Data Management encompasses the complete discipline. Neither provides the specific “directive” definition stated in the question.

Reference Topics: DAMA-DMBOK2 Chapter 3 — Data Policies; Governance Principles; Standards and Procedures; Chapter 13 — Data Quality Governance; Metadata and Master Data interactions.

===============

A process can be statistically stable yet still perform at an unacceptable quality level. Statistical control means that variation is predictable within the established process; it does not mean that the average level of defects satisfies business expectations.

For example, a process may consistently produce a 3% defect rate with very little month-to-month variation. If the business requirement is below 0.5%, the process is stable but incapable of meeting the desired quality level without improvement.

The appropriate response is therefore deliberate process improvement aimed at changing the underlying process and reducing its baseline defect rate. This differs from reacting to random individual observations within normal control limits.

After improvement, new performance data should be collected and control parameters reevaluated once the revised process reaches a stable state.

DAMA's revised Chapter 13 explicitly maps Shewhart and Deming improvement-cycle stages to Data Quality processes and preserves Statistical Process Control as supporting material.

The central principle is that control and capability are different questions: control asks whether the process is stable; business quality requirements determine whether that stable performance is good enough.

Reference Topics: DAMA-DMBOK2 Chapter 13 — Statistical Process Control; Process Stability; Continuous Improvement; Shewhart/Deming Cycle; Thresholds.

===============